The idea that correlation does not imply causation is a fundamental caveat in epidemiological research. A classic example involves a hypothetical link between ice cream sales and drownings. Rather than increased ice cream consumption causing more people to drown, it is likely that a third variable, summer weather, fuels ice cream cravings and swimming, and thus opportunities. drowning.
But what about correlations with genes? How can researchers be sure that a particular trait or disease is really genetically linked and not caused by something else?
We are statistical geneticists who study the genetic and non-genetic factors that influence human variation. In our recently published research, we found that the genetic links between traits found in many studies may not be linked by genes at all. Instead, many are a result of how people mate.
Genome-wide association studies try to link genes to traits
Since the genes you inherit from your parents remain unchanged throughout your life, with a rare exception, it makes sense to assume that there is a causal relationship between certain traits you have and your genetics.
This logic is the basis for it genome-wide association studies, or GWAS. These studies collect DNA from many people to identify positions in the genome that may be correlated with a trait of interest. For example, if you use certain forms of the BRCA1 and BRCA2 genesyou may have an increased risk of certain types of cancer.
Likewise, there may be gene variants that play a role in whether or not you have schizophrenia. The hope is to learn something about the complex mechanisms that link variation at the molecular level to individual differences. With a better understanding of the genetic basis of different traits, scientists would be better able to determine risk factors for related diseases.
Researchers have walked thousands of GWAS to date, identifying genetic variants associated with numerous diseases and disease-related traits. In many cases, researchers have identified genetic variants that affect more than one trait. This form of biological overlap, in which the same genes are believed to influence different apparently unrelated traits, is known as pleiotropy. For example, certain variants of the PAH gene can have different separate effectsincluding changing skin pigmentation and causing seizures.
One way scientists judge pleiotropy is through genetic correlation analysis. Here, geneticists investigate whether the genes associated with a particular trait are associated with other traits or diseases by statistically analyzing large samples of genetic data. Over the past decade, genetic correlation analysis has become the primary method for assessing potential pleiotropy in a variety of fields internal medicine, social Sciences and psychiatry.
Scientists use the findings of genetic correlation analyzes to uncover the possible shared causes of these traits. For example genes associated with bipolar disorder also predict anxiety disorders, perhaps the two conditions may partially involve some of the same neural circuits or respond to similar treatments.
Assortative mating and genetic correlation
However, just because a gene is correlated with two or more traits doesn’t necessarily mean it causes them.
Virtually all statistical methods researchers commonly use to assess genetic correlations assume mating is random. That is, they assume that potential mating partners decide who they will have children with based on a roll of the dice. In reality, there are probably many factors that influence who mates with whom. The simplest example of this is geography: people who live in different parts of the world are less likely to end up together than people who live close by.
We wanted to know to what extent the assumption of random mating affects the accuracy of genetic correlation analyses. In particular, we focused on the possible disruptive effects of assortment mating, or how people tend to mate with those who share the same characteristics with them. Assortative mating is a widely documented phenomenon seen in a wide variety of traits, interests, measures, and social factors, including height, education and psychiatric disorders.
In our study we examined cross-trait assortative mating, in which people with one trait (e.g., being tall) tend to mate with people with an entirely different trait (e.g., being rich). From our database of 413,980 mate pairs in the United Kingdom and Denmark, we found evidence of cross-trait assortative mating for many traits. other characteristics, including height, smoking status and risk of various diseases.
We found that considering the similarities between partners could strongly predict which traits would be considered genetically linked. In other words, based on the number of traits a partner pair shared alone, we were able to identify about 75% of the presumed genetic links between these traits – all without taking any DNA samples.
Genetic correlation does not imply causation
Cross-trait assortative mating forms the genome. If people with one inherited trait tend to mate with people with another inherited trait, then these two different traits will become genetically correlated in subsequent generations. This will happen regardless of whether these traits are really genetically linked.
Cross-trait assortative mating means that the genes you inherit from one parent will be correlated with the genes you inherit from the other. How people mate is not random and violates the main assumption behind genetic correlation analyses. This increases the genetic association between traits that are not actually linked by genes.
Recent studies confirm our findings. Earlier this year, researchers calculated genetic correlations using a method that examines the association between the traits and genes of siblings. The genetic links between traits affected by assortative mating between traits were significantly attenuated.
But without taking into account cross-trait assortative mating, it can be misleading to use genetic correlation estimates to study the biological pathways that cause disease. Genes that influence only one trait appear to influence several different conditions. For example, a genetic test designed to assess risk for one disease may falsely detect vulnerability to a large number of unrelated conditions.
The ability to measure variation between individuals at the genetic and molecular level is truly an achievement of modern science. However, genetic epidemiology is still an observational endeavor, subject to the same caveats and challenges that other forms of non-experimental research face. While our findings do not exclude all genetic epidemiological research, it will be essential to understand what genetic studies really measure in order to translate research findings into new ways to treat and assess disease.